Re: Names and addresses

I think that the distinction of being a name or an address is not a 
property of any particular label-token, but a consequence of how it is 
used.  I think it's fairly clear that any (unambiguous) address *can be 
used* as a name.  I also think that, given an appropriate infrastructure, 
any name can be used as an address.  Whatever the standards community may 
decree, users will not feel constrained not to use the labels we provide -- 
URIs -- in any way that seems convenient to their purpose.

So while naming and addressing may be separate problems, I don't think we 
can insist that they use disjoint sets of tokens.

#g
--

At 10:13 16/10/03 +0200, Gustaf Liljegren wrote:

>I don't like to interrupt the standardization going on in this group, but
>I've grown increasingly frustrated that people with otherwise very
>respectable opinions seem to "outgrow" the idea that names and addresses
>should be kept separate. Why?
>
>I like the distinction between names and addresses. While URIs are
>generally considered opaque, I think one should at least be able to see on
>a scheme name what is supposed to happen when you activate/use the URI.
>
>Consequently, I don't like seeing "http" being used for other things than
>addresses to resources that are retrievable from a web (HTTP) server. I'm
>using addresses for things that are network retrievable, and names for
>things that are not. From my point of view, there is never a need to
>resolve a name.
>
>Names are useful to make sure two programs are talking about the same
>things. Addresses are useful to fetch resources. I think these are two
>separate problems. Names have good use in ontologies. Fetching resourses is
>an entirely different matter.
>
>My world may be small, but this view is crystal clear to me. What are the
>counter-arguments?
>
>Gustaf

------------
Graham Klyne
For email:
http://www.ninebynine.org/#Contact

Received on Thursday, 16 October 2003 05:14:19 UTC